JLAB-TN-05-048
Radiological Sampling and Gamma Scans Aboard the N.S. Savannah
Conducted for the U.S. Maritime Administration
April, 2005
Keith Welch, Erik Abkemeier, Zachary Edwards
Jefferson Lab, 12000 Jefferson Ave.Newport News, VA23606
Background
Thomas Jefferson National Accelerator Facility (Jefferson Lab) has entered into an agreement with the U.S. Maritime Administration (MARAD) to provide support for the Savannah Emergency Response Assessment Team (SERAT) efforts related to N.S. Savannah in the event of an incident that might have radiological implications (Reimbursable Agreement #MA-5-A04, to fund Work for Others Project SURA 2004W007). The lab’s role is advisory, related to the health physics concerns associated with initial response activities in the event of an emergency.
Commensurate with its role of health physics support for SERAT efforts, Jefferson Lab conducted a series of measurements to confirm the primary nuclides of concern remaining in the reactor systems of the Savannah. This report details the findings of the measurements.
Acknowledgements
Jefferson Lab was afforded access to the primary plant areas of Savannah in conjunction with work being conducted by WPISM (WPI). WPI performed radiological and environmental assessments aboard Savannah which required opening all reactor-related spaces. The assessment included a breach of the primary reactor cooling water system in which samples were taken by WPI. WPI provided various samples to Jefferson Lab for analysis. Jefferson Lab acknowledges the kind assistance of WPI in obtaining data for this report.
Introduction
The purpose of conducting the measurements described in this report is to obtain a measurement-based estimate ofthe quantity and distribution of radionuclides in reactor primary systems aboard Savannah. The specific focus of this assessment is radioactivity in residual liquids and transferable contamination that may be subject to a spill or spread in the event of damage to the ship or flooding of compartments containing reactor systems. In the event of an emergency, knowledge of the reactor plant nuclide inventory is important in facilitating emergency response efforts in which Jefferson Lab may be involved. Previous calculations have been conducted to estimate volumetric nuclide content in the reactor vessel(1). This report does not address volumetric activation, but rather the distribution of internal surface contamination and contaminated liquids within piping and components of the primary system. The quantity of radioactivity deposited in the systemas contamination is very small compared to the total activity in reactor vessel components.
Internal surface contamination contenthas been estimated previously, but the present assessment effort provided a rare opportunity to re-evaluate radioactivity estimates based on a combination of measurement methods.
General Approach
Two methods were used to assess the residual radioactivity in the reactor systems. One method was to analyze samples from within the primary system. The second method was by direct scans with portable gamma spectroscopy equipment.
Samples (smears and liquid) taken from the primary system and within the reactor containment were quantitatively analyzed by high-resolution gamma spectroscopy at the Jefferson Lab Radiation Control Group (RCG) radioanalytical lab for presence and amounts of gamma-emitting radionuclides.
The direct gamma scan survey is qualitative in nature – its goal being to gather “snapshots” of the radiation field around various system components to further enhance the understanding of primary system nuclide content.
Equipment and Measurement Techniques
Measurements of sample media were made with a Canberra Industries ultra-high purity, coaxial germanium detector (relative efficiency, ~20%) with associated NIM electronics, operated via the Canberra Genie® software package. The system is energy and efficiency calibrated for a number of sample geometries annually and receives daily quality assurance checks according to Jefferson Lab RCG procedures. Jefferson Lab also participates in the Department Of Energy (DOE) Mixed Analyte Performance Evaluation Program (MAPEP) for measurement quality assurance.
Onboard gamma spectra were collected primarily with a Berkeley Nucleonics Corporation model SAM-935® portable surveillance and measurement system, consisting of a 3x3 NaI(Tl) detector coupled to the base unit electronics. The collected spectra can be analyzed with built-in software or uploaded to a PC for analysis using third-party software. A few spectra were also collected with a portable high-resolution germanium detector coupled to a Canberra Inspector® electronics package and analyzed using the Genie® software. This system proved to be difficult to manage in the shipboard environment due to its bulkiness and required a lengthy stabilization period each time the detector was shut down for movement and subsequently restarted.
Energy calibration of the SAM-935 is initially conducted by the factory using a multi-nuclide source. The calibration coefficients are stored in the firmware of the instrument. Field adjustment/drift-correction of the energy calibration is done with an automated calibration routine using a small Cs-137 check source. This routine can be conducted repeatedly at the user’s discretion. In addition, to enhance the accuracy the field measurements, some spectra were collected with reference sources present. The reference sources provided gamma rays of known energy, which can be used for a posteriori energy calibration corrections.
Nuclide identification from the SAM spectra was conducted using on-board analysis routines. Some of the spectra were also analyzed using a third-party program, PGT Quantum® gamma analysis software. This was done to conduct manual energy calibration corrections that allowed better photopeak identification when a peak could not be confidently identified by the SAM. Quantum also contains a superior nuclide library.
Energy calibration of the portable high resolution system was initially conducted at Jefferson Lab, with manual fine adjustments made in the field using reference peaks from small sources and known nuclides in the sampled spectrum.
Scope and Limitations
The direct survey is limited to those nuclides which decay with gamma emissions between approximately 30 keV and 3 MeV. Locations for measurement were chosen with the intent to monitor a reasonable cross-section of systems that contain radioactivity. Consideration had to be given to ambient radiation intensity such that the monitoring system could acquire spectra without encountering detector saturation problems (ambient radiation fields above about 1 mR/h cause significant detector dead-time), as well as the physical constraints of manipulating the detector and associated equipment within the spaces aboard the ship and protecting the equipment from potential radioactive contamination. Several locations within and outside the primary containment were monitored. Since these measurements were made in a “general area” radiation field involving complex source geometries, quantitative results regarding the concentration of radioactive material are not possible. However, gamma energy peaks provide qualitative verification of the presence and distribution of the most predominant gamma-emitting nuclides.
A limitation inherent in all the area scans is that the spectra include contributions from all sources in the vicinity of the item being monitored. One cannot determine conclusively that the activity indicated is attributable exclusively to the item of interest. Another limitation in assessing the contents of components is the self-shielding of the radiation by the components themselves.
Analysis of samples from the primary system provides the best opportunity to determine what nuclides might be present in the event of a spill from the system. The gamma analysis system usedfor sample counting has a functional energy range of about 5 – 2000 keV. Detector response extends below 5 keV (making detection of Fe55 possible in principle), but sample configuration and self-shielding probably prevent detection of photons below about 5 –7 keV.
One goal of the WPI assessment team was to investigatethe existence and quantity of water in the primary system beyond the reactor vessel. Steam generator hot-leg access was performed for this purpose. It was discovered that a significant quantity of water was present in the generators and lower hot-leg piping. Smear and water samples were obtained from inside the steam generators. An estimate of the total contamination inventory is made based on samples from the starboard steam generator. Also analyzed were smear samples from the primary containment enclosure that showed positive results during gross alpha/beta counting.
General Findings
Co60 was expected to be the most widespread nuclide in the primary system. This expectation was confirmed in the measurements taken. All the area monitoring spectra taken around primary systems indicated Co60 activity. Most monitored locations also indicated the presence of Cs137 (this may have some practical implications, as is discussed below). A photopeak present in some of the spectra at approximately 75 keV is attributed to lead fluorescence X-rays (Kα - 72.8 keV, Kβ - 74.9 keV), as significant quantities of lead shielding are present around the reactor vessel and in other monitored areas.
The WPI assessment team found very little surface contamination external to primary system piping and components. A few smear samples from reactor spaces showed a combination of Co60 and Cs137. In one case, only Cs137 was present. This is reasonable given the low activity in that area and the ratio of Cs137 to Co60 on the other smears (see detailed findings). It might also be surmised that the presence of the contamination is due to past spills of system coolant or ion-exchange media, rather than the dry release of crud from piping internals. This deduction is discussed further below.
Detailed Findings
Samples from inside the primary system showed the following characteristics. The steam generator water sample contained Cs137 almost exclusively (Cs137 concentration was about 1000 times greater than Co60), but contamination on interior surfaces of the steam generator was found to contain only Co60. This is undoubtedly a result of the chemical form of the contaminants. Co60 is usually found as an insoluble oxide, and tends to deposit on surfaces of reactor systems (forming the common “crud” deposits found in all reactors), whereas Cs137 is present as a very soluble oxide or hydroxide.
A spill of the coolant would be likely to spread both Co60 and Cs137, as the Co60 is easily removable and would be flushed from surfaces by any significant movement of the water (hence the speculation above that contamination on surfaces in the reactor compartment may be the result of past liquid (or ion-exchange media) spills). A spill to the environment (i.e. into the James River) would probably behave similarly with respect to the distribution of these nuclides. The Cs137 would likely remain dissolved in the river water, whereas insoluble components would eventually find their way into sediment.
The tables below summarize the area monitoring and sample analysis results. The area scans performed with the SAM 935 contain exposure rate estimates associated with the identified nuclides. This is a calculation made by the SAM using an algorithm that converts counts in a photopeak to an energy-corrected exposure rate. The exposure rate indication provides a reasonably accurate relative intensity measurement.
Area Monitoring Results
Scan Ref # / Location / Component or System / Nuclides and Exposure Rate(uR/h) / Cs137/Co60
Exp. Rate Ratio / Notes
M1 / Hold Deck, Port passageway / 4” Piping below deck level / Cs137 / 0.95 / 0.03
Co60 / 30
M2 / HoldDeckPort passageway / Aft end of passage, effluent piping under deck / Cs137 / 0.08 / 0.05
Co60 / 1.6
M3 / HoldDeckPort passageway / Small diameter pipe behind cage chg. pmp. buffer seal system / Cs137 / 1.4 / 0.24
Co60 / 5.9
M4 / Port Stabilizer Room / 6” piping from charging pump buffer seal system / Cs137 / 12.1 / 0.56 / 1
Co60 / 21.6
M5 / Prim. Containment Upper level / Primary coolant line interface to reactor vessel Fwd / Co60 / 581 / N/A
M6 / Prim. Containment Upper Level / Primary coolant line interface to reactor vessel Aft / Co60 / 564 / N/A
M7 / Prim. Containment Upper Level / Upper pressurizer head, Port / Co60 / 183 / N/A
M8 / Prim. Containment Upper Level / Fwd upper Regen/non-regen heat exchanger / Cs137 / 6 / 0.025
Co60 / 242
M9 / Prim. Containment 2nd Level / Crossover line from upper to lower Regen/non-regen HtXchr / Cs137 / 8.5 / 0.062
Co60 / 137
M10 / Prim. Containment 3rd Level / Main pressurizer leg to prim. coolant line, just under preszr. / Cs137 / 6.8 / 0.044
Co60 / 156
M11 / Prim. Containment 4th Level / Check valve adjacent to Fwd primary coolant line near vessel / Cs137 / 19 / 0.053
Co60 / 360
M12 / Prim. Containment 4th Level / Reactor vessel (shield tank wall) Fwd, just Stb. of center / Cs137 / 11 / 0.023
Co60 / 479
M13 / Prim. Containment 1st Level / Rx ventilation plenum duct, Stbd / Cs137 / 18 / 0.21 / 2
Co60 / 84.5
M14 / Cold Chem lab Upper Level / Rx ventilation duct / Cs137 / 0.043 / 0.17 / 2
Co60 / 0.25
M15 / Cold Chem lab
Lower Level / Primary sample sink, sample bulb inside sink hood / Cs137 / 3.6 / 0.015
Co60 / 242
M16 / Port Charge Pump Room / Between pumps at Aft bulkhead / Cs137 / -- / 0.096 / 3
Co60 / --
Table 1: Area monitor scan results
1 – Exposure rates estimated
2 – Measurement on ventilation ducts
3 – Measured with high resolution Ge detector. Ratio taken from peak area data.
Sample Analysis Results
Ref # / Location/Component / Sample Type / Nuclides / ActivityS1 / Stbd. Steam generator tube sheet / Smear / Co60 / 144,300 dpm/100cm2
S2 / Stbd. Steam generator interior (average)1 / Smear / Co60 / 22,000 dpm/100cm2
W1 / Stdb. Steam generator water / 100 ml water / Cs137 / 1.04E-3 µCi/ml
Co60 / 1.45E-6 µCi/ml
S3 / Reactor 3rd Level Fwd at Przr. (highest)2 / Smear / Cs137 / 1200 dpm/100cm2
Co60 / 250 dpm/100cm2
S4 / Reactor 1st Level Fwd Rx head (average)3 / Smear / Cs137 / 350 dpm/100cm2
Table 2 – Laboratory gamma analysis of samples
1 Average of 4 smears, excludes tube sheet
2 Composite count of 6 smears, all activity attributed to one smear
3 Composite count of 5 smears, activity averaged over the total
Calculation of Total Contamination Inventory
The total contamination inventory for the primary system was estimated based on the sample data. The contamination inventory is broken into two parts; (1) internal surface contamination, and (2) contamination entrained in residual coolant.
Surface Contamination
The surface contamination estimate begins with an assessment of the steam generator contamination content. Published industry data(2) indicate that in PWRs, the majority of coolant-borne corrosion/fission products that are not removed by the chemical volume and control system (CVCS) are deposited in the steam generators. For a reference PWR*, the generators contain about 85% of the total deposited activation product inventory. The balance of the activity is distributed in various other components based on relative surface area and deposition characteristics of the system/component.
Steam generator activity content was estimated based on the highest contamination level found in the starboard generator. Assumptions for the calculation are as follows.
- The only nuclide of concern for surface contamination is Co60
- Smears were taken over a 100cm2 area
- The removal factor for smears is assumed to be 0.1
Steam Generator Dimensional Estimates
-Tube diameter: 0.5” (1.27 cm)
-Average tube length: 30’ (900 cm)
-Number of tubes: 2000
-Shell ID: 100 cm
-Total plenum length: 100 cm
* The Reference PWR in the literature was the Trojan Nuclear Plant. Distribution of radioactivity in three other PWRs was evaluatedand reported in Ref. 2. The percentage of radioactivity deposited in steam generators was similar in each case.
Tube surface area: 2π(0.635)(900)(2000) = 7.18E6 cm2
Total tube sheet area: 2[π(50)2 – π(0.635)2(2000)] = 1.06E4 cm2
Plenum area: 2π(50)(100) = 3.14E4 cm2
Internal surface area of one steam generator: 7.18E6 + 1.06E4 + 3.14E4 = 7.2E6 cm2
Total activity in one generator in curies is calculated as follows:
144,300 dpm x 7.2E6 cm2 = 0.0468 Ci or, 93.6 mCi for both steam generators
0.1 x 100 cm2 x 2.22E12
Adjusting for reactor/steam generator surface area ratios and unit layout (2-loop vs. 4-loop), activity distribution assignmentswere made based on the reference PWR. Associated activity levels were calculated and are summarized in Table 3.
System / Activity Distribution (%) / Total Activity (Ci)Reactor Vessel and Internals / 5 / 0.0054*
Steam Generators / 87 / 0.0936
RCS1 Piping / 3 / 0.0032
Non-RCS Piping / 2.3 / 0.0025
Pressurizer / 0.2 / 0.0002
Other / 2.5 / 0.0027
Totals / 100 / 0.108
Table 3 – Total Surface Contamination Inventory
*Excludes volumetrically distributed activation products in the reactor vessel
1RCS = Reactor Cooling System (main cooling loops)
Contamination in Residual Coolant
Using visual indications from the steam generator coolant content, the estimated volume of water in the primary system is calculated below, with the associated total radioactivity.
Volume of generator primary side: π(0.635)2(900)(2000) + π(50)2(100) = 3.1E6 cc (ml)
In addition, a portion of the RCS hot and cold legs run horizontally into and out of the generator. The total length of this piping is estimated to be about 26 feet (780 cm) for each loop. The piping diameter is estimated at 18” (45cm).
Volume of horizontal piping: π(22.5)2(780) = 1.2E6 ml
Total volume of contiguous horizontal coolant envelope (1 loop): 1.2E6 + 3.1E6 = 4.6E6 ml
The water level in the starboard generator was observed to be about halfway up the generator tube sheet, the port generator was reported to be about one third full. For this estimate, both will be considered half full.
Total water volume in horizontal legs: 4.6E6 (2) = 4.6E6 ml (~1200 gal)
2
It has been estimated by others that about 1100 gallons of water reside in the lower reactor head. We estimate another 200 gallons is distributed around the balance of the reactor systems (this is based partly on the observation discussed below regarding location of liquid via the presence of Cs137). This brings the total volume to 2500 gallons (9.5E6 ml). Assuming the activity in the water is uniform through the plant, and represented by the activity in the steam generator, the total activity is:
Cs137 – (1.04E-3µCi/ml)(9.5E6 ml) = 9840 µCi
Co60 – (1.45E-6 µCi/ml)(9.5E6 ml) = 14 µCi
Additional Observations and Some Speculation
The observed distribution of Co60 and Cs137 might serve as an indicator of the presence of liquid within various systems and components. If the same physical separation of nuclides found in the steam generator is assumed to exist throughout the system, one could use the presence of Cs137 in an area scan of primary piping as an indicator of liquid in the component in question. If only Co60 is present, it may be an indication that the piping or component is internally dry or contains little liquid.
The results of the area scans taken qualitatively support this idea. For instance, no Cs137 was seen in scans of the upper main coolant lines at their interface to the reactor vessel. By comparison, all the scans of the lower level reactor compartment (containing the primary side of the steam generators and other low-point piping) show Cs137. Although not conclusive, this is consistent with the hypothesis that dry piping contains little or no Cs137 contamination. The ratio of Cs137 to Co60 activity was found to be highest near piping outside the primary containment in the lowest levels of the ship (eg. piping in the Hold level, Stabilizer Room lower level and Charge Pump Room). Table 1 includes these ratios for information purposes.